Internal
Container-to-host internal routes
- Fetch run history
- Recall archive memories
- Fetch a fresh access token for an OAuth model provider connection
- Force a refresh of the access token for an OAuth model provider connection
- Fetch live credentials + HTTP delivery plans for an active integration
- Force-refresh OAuth2 credentials for an active integration
- End a connection's upstream-rejection streak
- Fetch the AFPS bundle bytes for a referenced mcp-server package
- Run inference (openai-completions) — metered by the platform LLM proxy
- Run inference (openai-responses) — metered by the platform LLM proxy
- Run inference (anthropic-messages) — metered by the platform LLM proxy
- Run inference (mistral-conversations) — metered by the platform LLM proxy